LG Electronics is building a facility in Seoul where robots will learn household and factory tasks using real and computer-generated data.
The Data Factory, located at LG’s Yangjae R&D Campus, is expected to be fully operational by the end of 2026. The four-floor facility will cover 10,000 square meters and house several hundred robots.
LG’s CLOiD home robots will repeatedly perform cleaning tasks in a space designed like a home. Other robots will practice moving, stacking, and assembling parts in an area modeled after LG’s washing machine plant in Tennessee.
LG CNS will use robots in logistics automation, while LG Innotek will train robotic hands in a separate area.
The facility will use NVIDIA Omniverse libraries, NVIDIA Cosmos open-world models, and the NVIDIA Isaac robotics development platform. These tools will help LG create additional computer-generated training data and use it to improve its robots.
By the end of 2026, LG expects to collect or generate 100,000 hours of training data, equal to about 12 years. The company plans to use the data to improve its Robot Foundation Model, which supports the performance of humanoid robots.
The project follows an agreement signed by LG Group and NVIDIA on Aug. 13 at NVIDIA’s headquarters in Santa Clara, California. The companies agreed to work together on future business projects.
LG has made robotics a larger part of its business in 2026. It recently formed a Robotics Business Center that reports directly to the CEO and oversees robotics projects across the company.
“Through the synergy built on ‘One LG’, bringing together core capabilities across the Group, and strategic collaboration with global partners, we will secure our competitiveness in physical AI and become a comprehensive robotics solutions provider,” said Lyu Jae-cheol, CEO of LG Electronics.
